r&d for safety, codes and standards: hydrogen behaviorthe needs of the project if the intent is...
TRANSCRIPT
R&D for Safety, Codes and Standards:Hydrogen BehaviorProject ID: SCS010
Ethan S. HechtWith contributions from: Pratikash Panda, Anthony McDaniel, Radoslav Bosinoski
Sandia National Laboratories
Sandia National Laboratories is a multi-mission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525. SAND NO. SAND2017-4818 C
This presentation does not contain any proprietary, confidential, or otherwise restricted information
June 6, 2017
Overview
BarriersA. Safety Data and Information:
Limited Access and AvailabilityG. Insufficient technical data to revise
standards
Timeline• Project start date: Oct. 2003• Project end date: Sept. 2017*
* Project continuation and direction determined by DOE annually
Budget• FY16 DOE Funding: $510 k• Planned FY17 DOE Funding: $500 k• Partner funding:
• $175k committed stakeholder funds (CaFCP Auto OEM Group, Linde, Shell)
Partners• Stakeholder CRADA
• Bki (contractor for California Fuel Cell Partnership)
• Fire Protection Research Foundation (research affiliate of NFPA)
• Industry & Research• NFPA 2 code committee• HySAFE
2
RelevanceObjectives: • Perform R&D to provide the science & engineering basis for the release,
ignition, and combustion behavior of hydrogen across its range of use (including high pressure and cryogenic)
• Develop models and tools to facilitate the assessment of the safety (risk) of H2 systems and enable use of that information for revising RCS and permitting stations
Barrier from 2015 SCS MYRDD Goal
A. Safety Data and Information: Limited Access and Availability
Build validated H2 behavior physics models that enable industry-led C&S revision and Quantitative Risk Assessment
G. Insufficient technical data to revise standards
Perform experiments to address targeted gaps in the understanding of H2 behavior physics
3
Relevance: Current separation distances for liquid hydrogen are based on consensus, not science• Previous work by this group led to science-based, reduced, gaseous H2 separation
distances• Higher energy density of liquid hydrogen over compressed H2 makes it more
economically favorable for larger fueling stations• Even with credits for insulation and fire-rated barrier wall 75 ft. offset to building
intakes and parking make footprint large
4
Approach (Sandia H2 SCS): Coordinated activities that facilitate deployment of hydrogen technologies• Hydrogen Behavior (this project, SCS010)
– Develop and validate scientific models toaccurately predict hazards and harm fromliquid releases, flames, etc.
• Quantitative Risk Assessment, tools R&D(SCS011)– Develop integrated methods and algorithms
enabling consistent, traceable, and rigorousQRA (Quantitative Risk Assessment) for H2facilities and vehicles
• Enable Hydrogen Infrastructure throughScience-based Codes and Standards (SCS025)– Apply QRA and behavior models to real
problems in hydrogen infrastructure andemerging technology
5
Approach: Develop and execute experiments to enable predictive modeling across H2’s range of useIssue: Cryogenic H2 releases have been outdoors and/or instrumented with low fidelity sensors, with experimental uncertainty too high for model validation• SNL Approach (FY17 goals):
– Perform lab-scale cryogenic hydrogen dispersion validation experiments with precise control of boundary conditions and high-fidelity imaging diagnostics (described in next slides)
– Design large indoor experiments to enable the modeling of pooling and vaporization of liquid hydrogen (described in remaining challenges)
Issue: Low fidelity sensors and very specific vent geometries in previous accumulation experiments• SNL Approach (FY17 goal): Design an experimental platform with
flexible vent/source geometry and high-fidelity, accurate concentration measurements (task to begin in Q4)
6
Approach (FY17 Hydrogen Behavior): Validate model for cryogenic hydrogen dispersion
• Previously developed model requires validation data– Several model parameters based on empirical data– Data only from warm hydrogen or other warm gases– Are more physics required?
• Use experimental platform commissioned in FY16 to generate cryogenichydrogen releases
Milestone (FY17Q2): Measure dispersion field of < 40 K hydrogen (milestonecomplete) and incorporate validated model into HyRAM (FY18 milestone)
7
Accomplishment: Developed and implemented Raman imaging technique to measure cryogenic plumes
• Conventional Rayleigh signaloverwhelmed by Miescattering off of condensedwater vapor in jet
• Filtered Rayleigh hadinsufficient Mie scatteringlight suppression (OD≈3)
• Raman scattering enableshigher optical density filters
– 10 nm FWHM bandpassfilters at wavelengths ofinterest
– OD of 12 @ all wavelengths– OD of 18 @ 532 nm
Enables simultaneous measurement of concentration and temperature in 2D
8
Accomplishment: Measured concentration and temperature fields of cryogenic hydrogen releases
Tnoz[K]
Pnoz[barabs]
d[mm]
Tthroat[K] nhts
58 2 1 43.5 456 3 1 41.9 453 4 1 39.6 450 5 1 37.4 561 2 1.25 45.7 651 2.5 1.25 38.2 251 3 1.25 38.2 655 3.5 1.25 41.2 354 4 1.25 40.4 2
Two-dimensional images are superior to centerline only measurementsfor model validation
9
Accomplishment: Measured temperatures as low as 85 K, 38 mm from the release point
Control of pressure, and measurement of temperature at nozzle enables data to be used for model validation
10
Accomplishment: When normalized properly, centerline and half-width decay rates scale linearly
• Literature inversemass-fraction decayrate: 0.21–0.271• Literature mass-
fraction half-widthspreading rate: 0.1–0.11 mm
First ever measurements of temperature needed for model energy balance11
Accomplishment: Radial profiles are self-similar, but wider than literature data of warm releases
Different profiles implies different modeling parameters from warm releases12
Accomplishment: Radial profiles of temperature are also self-similar, and wider than mass fraction
Data can be used to validate relative spreading ratio in model13
Progress: Initial comparisons to model show slower centerline mole-fraction decay rate than predictions
Expansion model and/or entrainment rate may need to be adjusted14
Progress: Initial comparisons to model show good agreement in terms of temperature
Improved data analysis may yield reduced noise15
Response to last year’s Reviewer’s comments
• The project needs more emphasis on adding liquid hydrogen capability. It would be helpful if the project could address releases in a container that did not trap hydrogen under a roof, thereby limiting the concentration.
– We recognize the ability to predictively model liquid hydrogen releases as being a critical barrier to code revisions; this work focuses directly on model validation. We are close to having a valid model for some aspects of liquid hydrogen stations. We are also planning some enclosure experiments to improve our simulations of scenarios with walls.
• The proposed work to evaluate cold/liquid releases does not appear adequate to meet the needs of the project if the intent is to address large-scale releases.
– The large-scale releases being planned (described in the following slides) will address the shortcomings of the lab-scale experiments, enabling modeling of larger releases.
• The project should add benchmarking against other fuels.– We are planning on running similar (cryogenic release) experiments with methane
(as a surrogate for natural gas), with funding from the vehicle technologies office.
Note: While this work was presented during the 2016 AMR, it was not independently reviewed. Since there are such close ties to SCS-011: Hydrogen behavior and Quantitative Risk Assessment, some relevant comments were taken from there.
16
Collaborations have enabled this research and expanded impact
H2 behavior (SCS010) collaborations• CRADA with BKi to fund experiments
– Commitments from Shell, Linde, CaFCP Auto OEM group
– inquires out to other industry organizations and local government agencies
• NFPA 2 Technical Code Committee– Regular attendance with expert
advisory role• HySafe
– Panelist for HySAFE Research Priority workshop on Hydrogen Safety
Expanded impact through HyRAM (SCS011) and C&S participation (SCS025)• HyRAM users – including ITM Power,
Paul Scherrer Inst., ZCES, AVT, …• Gexcon - Technical exchanges on
validation activities for physics models, integration of safety methodology approaches; In-kind support - provided FLACS research license
• PNNL - Technical exchanges on PBD;QRA; Hydrogen Safety Panel
• NREL - Technical exchanges on PBD; QRA• HySafe - Technical exchanges on safety
methodology; QRA toolkits • ISO TC197 WG24- SNL co-leads sub-
team on safety methodology • IEA HIA Task 37 -SNL leads sub-task on
Safety Integration Toolkits; • H2USA - Various working groups17
Remaining challenges: Phenomena from large-scale releases are not well understoodNeed experiments to characterize:• Pooling• Evaporation from LH2 pools
Planning underway for experiments at Sandia (Albuquerque) facilities:• Thermal test complex
– Flame cell• Up to 3m diameter pool• 18.3 m dia. x 12.2 m high• Well characterized conditions for model validation
– Crosswind test facility• Dispersion in controlled crosswind• Single-direction flow• Well-characterized ambient conditions
• Severe Accident Phenomena/Analysis (Surtsey)– 100 m3 pressure vessel with 6 levels of instrumentation ports
18
Proposed future work• Remainder of FY17
– Complete Raman imaging characterization of cryogenic hydrogen releases– Conclude development/validation of the ColdPLUME model for predictive
dispersion modeling– Develop R&D plans for large-scale experiments– Develop research plans and design enclosure/accumulation experiments
• FY18– Integrate validated ColdPLUME model into publicly released HyRAM– Conduct large-scale release experiments to characterize hydrogen pooling,
evaporation, and interaction with atmosphere– Complete enclosure/accumulation experiments and develop predictive models of
risk for unintended releases of hydrogen in containers• Out years
– Simulate scenarios driving separation distances in NFPA 2 and enable the science-based revision of the liquid hydrogen separation distances in the 2022 version of NFPA 2
Any proposed future work is subject to change based on funding levels19
Summary• Relevance: Address lack of safety data, technical information relevant to
development of Codes & Standards.• Approach: Develop and validate scientific models to accurately predict hazards
and harm from liquid releases, flames, etc. Generate validation data where it is lacking.
• Technical Accomplishments:– Raman imaging used to characterize concentration and temperature profiles of
cryogenic hydrogen releases– Preliminary analysis suggests that mixing of cryogenic hydrogen jets is different
from warm hydrogen jets, requiring new empirical model parameters– Experiments on-going
• Push to even lower temperatures
• Future work:– Modify ColdPLUME model as validation data dictates– Implement ColdPLUME into HyRAM– Perform large-scale experiments and develop models for pooling and evaporation– Use models to advise NFPA 2 code committee on hazards and harm for high
priority scenarios (to inform 2022 edition of NFPA 2)20
TECHNICAL BACKUP SLIDES
21
A conceptual model needs to further validation• Conservation of mass,
momentum, species, energy• 5-zones:
– Zone 0: accelerating flow– Zone 1: underexpanded jet– Zone 2: initial entrainment and heating– Zone 3: flow establishment– Zone 4: self-similar, established flow
• 1-dimensional along streamline, can curve due to buoyancy
Houf & Winters, IJHE, 201322
(Air) icing at the nozzle likely improves mixing for temperatures < 50K
23
Ignition distance and radiant fraction were mapped out last FY
24
Two high priority scenarios identified by the NFPA 2 code committee are initially targeted for modeling Flow from vent of ultra-cold hydrogen (e.g. trailer venting excess pressure
after normal LH2 delivery or burst disk rupture)– Are vent stacks appropriately designed?– Separation distance from air intakes and overhead utilities– Vertical discharge, 3” diameter pipe, 20-140 psig
Release from pipe containing liquid H2 (e.g. leading from tank to vaporizer or vaporizer itself - caused by thermal cycles or ice falling from vaporizers)
– Requires ability to model flashing, pooling and evaporation from pools– Need to model concentration plume and heat flux from a subsequent fire – Horizontal discharge, ¾”-2” diameter pipe, 20-140 psig
25
We are running an experiment, releasing ultra-cold hydrogen in the laboratory
Accurate control/measurementof boundary conditions
26